地震资料的多准则速度分析方法

速度分析是地震数据处理中至关重要的一环,常规速度分析方法中,给定一系列相同间隔的速度进行扫描,以叠加能量或相似系数等作为速度分析的准则制作速度谱.速度谱的能量团通常在正确的动校正速度附近存在一定的发散,从而影响速度的拾取.通过对方差为准则的速度分析方法进行分析和研究,认为方差对同相轴的相干性比较敏感,因此,方差速度谱具有较高的精度.拾取方差速度谱中最小方差数值对应的扫描速度为动校正速度,然而实际资料处理时,由于精度太高,最小方差数值对应的区域太小,不利于速度的拾取.为此笔者提出了以方差倒数、叠加能量和相似系数三者的乘积为目标函数的多准则速度分析方法.噪声会对方差产生影响,然而噪声只是提高方差的整体数值,不会改变方差的相对大小关系,因此地震数据含有一定噪声时本文方法仍然适用.通过模型数据和实际资料的处理,验证了本方法的正确性和实用性.

The multi-criteria velocity analysis of seismic data

Velocity analysis is a key procedure of the seismic data processing,the conventional velocity analysis method scans with a series of given velocities with the same interval,and makes velocity spectrum based on the criteria of stacking power,semblance coefficient and so on.The energy groups of conventional velocity spectrum are usually divergent at the right NMO correction velocity,which affects the velocity pick up.A variance velocity method is studied and its velocity spectrum has higher precision since variance is relatively sensitive to the coherence of the events.The scanning velocity corresponding to the minimum values of the covariance velocity spectrum is picked up as the NMO correction velocity.However,due to the high precision,the area corresponding to the minimum variance value is too small in actual data processing,which is unfavorable for velocity pick up.In view of such a situation,the authors proposed a multi-criteria velocity analysis method regarding the production of variance,stacking power and semblance coefficient as the objective function.Noise will affect the variance by increasing its overall value but does not change its relative value relation.This method can also be applied to the seismic data with certain noise consequently.The authors have proved the correctness and effectiveness of the method by model data and actual data processing.